Terminal Iterative Learning Control for Nonaffine Nonlinear Systems with Nonrepetitive Uncertainties

In this paper, a terminal iterative learning control (TILC) is proposed for discrete-time nonaffine nonlinear systems with nonrepetitive uncertainties in the plant model, external disturbances, initial states, and reference trajectories. An extended terminal dynamic linearization (TIDL) method is utilized to transform the nonlinear system into a linear data model. A data-driven double dynamics analysis method is presented for TILC, by which we can demonstrate that although there exist nonrepetitive plant model uncertainties, the I/O data is bounded and the tracking error is robustly convergent. The developed analysis method is data-driven since no model information is utilized. Simulations verify the developed theoretical results.